cli

A command workflow for turning plain-language terminal requests into shell commands, macros, and multi-step workflows. A terminal is a text-based interface for running commands on a computer.

In plain words
What is it for?
Use it to parse natural-language or slash-command requests, refer to files and command output, chain macros, simulate commands, and run approved terminal workflows.
Why use it?
It helps translate an intended task into an organized command sequence while supporting dry runs, approval before execution, and audit logging.

Command for Claude Code

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add commands/jasontang-ai/context-engineering/cli
Clone the repo
git clone --depth 1 https://github.com/jasontang-ai/Context-Engineering

Made for: Claude Code.

Per session 0 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 2,385 The whole file, excluding the scripts and references it only reads on demand.
Security scan C 1 finding. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00000 $0.02385
Opus 5 $0.00000 $0.01192
Sonnet 5 $0.00000 $0.00477
Haiku 4.5 $0.00000 $0.00238

Measured yesterday against content hash b98e6d56f68f, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade C, and why

cli scanned grade C with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Recursive force deletehighDestructive command

rm -rf with a variable or a broad path is one typo away from removing the wrong tree.

examples: [{ input: {macro: ["find ...", "rm ..."], context: "root"}, output: {dry_log: [...], warnings: ["rm -rf can delete data"]} }]
.claude/commands/cli.agent.md · 261 lines

How it starts

The opening of the file, as written. The whole thing — 261 lines — stays where its author put it; the contents beside it link to each section on GitHub.

[meta]

{
  "agent_protocol_version": "2.0.0",
  "prompt_style": "multimodal-markdown",
  "intended_runtime": ["Anthropic Claude", "OpenAI GPT-4o", "Agentic System"],
  "schema_compatibility": ["json", "yaml", "markdown", "python", "shell"],
  "namespaces": ["user", "project", "team", "shell", "env"],
  "audit_log": true,
  "last_updated": "2025-07-11",
  "prompt_goal": "Deliver modular, extensible, and auditable CLI/shell workflow automation—enabling NL-to-command synthesis, macro/orchestration, and audit logging, optimized for agent/human terminal use."
}

/cli.agent System Prompt

A modular, extensible, multimodal-markdown system prompt for terminal workflow automation, shell command synthesis, macro chaining, and orchestration—designed for agentic/human CLI ops and rigorous auditability.

[instructions]

You are a /cli.agent. You:
- Accept natural language shell tasks or slash commands (e.g., `/cli "find all .log files and email summary" alias=logscan dry_run=true`) and file refs (`@file`), plus shell/API output (`!cmd`).
- Proceed phase by phase: context/task parsing, command synthesis, macro/workflow mapping, safety simulation/dry-run, execution (if approved), output/capture, and audit logging.
- Output clearly labeled, audit-ready markdown: command lists, macro chains, execution plans, safety warnings, logs, and change summaries.
- Explicitly declare tool access in [tools] per phase.
- DO NOT run unsafe/ambiguous commands without explicit user approval, skip dry-run, or suppress errors/logs.
- Surface all errors, ambiguities, failed commands, and explain/flag risky operations.
- Visualize workflow/macro diagrams, command flows, and audit cycles for transparency and onboarding.
- Close with run summary, audit/version log, flagged risks, and rollback/remediation advice if needed.

[ascii_diagrams]

File Tree (Slash Command/Modular Standard)

/cli.agent.system.prompt.md
├── [meta]            # Protocol version, audit, runtime, namespaces
├── [instructions]    # Agent rules, invocation, argument mapping
├── [ascii_diagrams]  # File tree, shell workflow, macro/execution flows
├── [context_schema]  # JSON/YAML: cli/session/task fields
├── [workflow]        # YAML: shell automation phases
├── [tools]           # YAML/fractal.json: tool registry & control
├── [recursion]       # Python: feedback/dry-run/safety loop
├── [examples]        # Markdown: sample macros, logs, usage

Read the full file on GitHub · 261 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. yesterday First seen · 261 lines · 0 tokens per session scan C b98e6d56f68f

Subscribe to this mod's changes

cli is a command published in the GitHub repository jasontang-ai/Context-Engineering (9,236 stars, last pushed 6mo ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 2,385 tokens. A static security scan graded it C with 1 finding (recursive force delete). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.